The present invention relates in general to telecommunication techniques. More particularly, the invention provides a system and method for low-cost multi-channel DWDM line card based on monolithically/hybrid integrated DWDM transmitter/receiver arrays. Merely by way of example, an application of the invention is a cost-competitive replacement of an EDFA, ROADM as well as a wavelength switch, but it should be recognized that the invention has a broader range of applicability.
Dense wavelength division multiplexing (DWDM), since its deployment in 1990s, has become a driving force for the rapid growth of various traffic in the long haul, regional as well as metro area networks. Recent convergence of video, voice and data, and the explosion of new applications such as video podcasting and peer-to-peer file sharing pose significant challenges to DWDM engineers to meet the dynamic wavelength demands for user services. For example, the traditional DWDM networks are based on EDFAs, DCMs and fixed OADMs and often unable to provision unplanned wavelength services. Supporting a new service in a current DWDM network often translates to the extended downtime of the network, increasing significantly service provider's operation expenses. To accommodate the changes due to service demands and traffic growth, it is often desirable for service providers to simplify their DWDM networks and move towards more efficient and flexible network managements while still maintaining high levels of reliability. Thus, it is desirable for new DWDM networks to meet basic requirements, such as simplicity, flexibility, robustness and bandwidth utilization efficiency.
Reconfigurable OADM (ROADM) is one of the promising optical solutions developed recently to meet the rapid increase of wavelength-service demands. With ROADM, service providers can dynamically provision the networks, adding new services and/or reallocating unused capacity. However, ROADM often has addressed only the flexibility aspect of the DWDM networks, leaving unresolved other issues such as simplicity, robustness and bandwidth utilization efficiency. In some cases, ROADM actually can complicate these issues. For example, in a ROADM-based ring network for metro applications, dynamic provisioning with ROADM often requires sophisticated optical power management algorithms with pre-knowledge of the new configuration to be able to re-evaluate parameters such as OSNR and mix-n-match penalty. Any change in network configuration with ROADM usually result in a performance change of every link in the network due to shared nature of EDFAs. Adding a new node can further complicate the ROADM-based networks and usually requires re-engineering because of, for example, inadequate OSNR.
On the other hand, recent development of integrated DWDM transmitter/receiver offers a potential solution to the next generation DWDM networks. For example, the arrayed DWDM LR transmitters/receivers may allow optical signals of different wavelengths to be first converted to electrical signals at every node regardless of their final destinations and reconverted back to optical signals, if not to be dropped, for transmission over fiber to the downstream node. This full regeneration approach in principle can eliminate EDFAs and DCMs and hence the optical power management, simplifying considerably the DWDM networks. In addition, with the use of a high-speed N×N electrical switch in the line card, one can configure the node to become effectively an enhanced ROADM node that offers signal regeneration and ROADM functions simultaneously. Furthermore, bandwidth efficiency and robustness (excluding fiber cut) often can be realized by using 1:N shared protection of transponder arrays. Studies show that with 1:12 protection re-generation using arrayed transponders can have better reliability than an EDFA.
But the integrated DWDM transmitter/receiver arrays, although promising from the performance point of view, are costly as a replacement for EDFAs and/or ROADM. It is usually not economically viable unless the cost of two such array units (which makes a re-generation/ROADM node) becomes less than that of an EDFA plus an ROADM. For example, the monolithic integration on InP can not meet this cost target. Even with recently proposed low-cost hybrid integration, the use of arrayed transponders to replace an EDFA and ROADM is still too costly to justify its deployment in many applications. For example, the associated electronic components such as FEC chips are the cost bottleneck.
Hence it is highly desirable to improve techniques for DWDM systems.